基于共振模型的操作风险相关性度量研究
|
摘要
对银行业来说，针对“操作风险”的研究已有近几十年的时间，其与信用风险、市场风险一起被视为商业银行面临的主要三大风险。分析操作风险的特征，可以发现其发生频率之高，目前操作风险损失几乎每天都会在商业银行的经营过程当中出现。其涉及范围之广，操作风险几乎涵盖了商业银行各个业务条线的运行。其计量难度之大，目前仍然没有一个模型能够完全准确的度量操作风险损失的大小。所以，关于操作风险的度量一直都是理论界关注的热点。2004年《巴塞尔新资本协议》（Basel ⅡAccord）首次对操作风险进行了严格的定义，并且提出了三种资本金计量方法：基本指标法、标准法和高级计量法。目前应用高级计量法度量操作风险是各大商业银行的趋势，并且现如今针对操作风险的研究也主要集中在高级计量法中，因为商业银行使用高级计量法可以根据自身的经营特征、内外部环境以及风险管理水平合理的估计面临的风险状况，更加精确的估计应计提的风险准备金，尽量减少因风险资本金计量偏差所导致的盈利能力下降或者是抵御风险能力不足的情况，从而为商业银行安全、稳健、快速的发展打下夯实的基础。
在操作风险的高级计量法当中，相关性的问题一直没有得到很好的解决。而相关性问题可以理解为是一种传染的效应，风险同时发生或者同时不发生。如果风险在很长时间之内同时不发生，那么商业银行可能会渐渐忽略此种风险的存在，逐步减少此项资本金的计提。但是，一旦风险爆发，很可能会严重威胁商业银行的生存与发展，造成无法弥补的损失。所以针对操作风险相关性的问题，本文首先在各个方面系统分析了操作风险的成因、内涵、特点、性质以及现有的各种度量方法，接着引入了操作风险相关性度量问题，指出目前使用线性相关关系度量相关性的缺陷，然后探讨了相关性问题研究的最新进展，给出两种相关性度量理论——Copula函数和Common Shock模型，并在对比分析两种相关性理论的基础之上，根据现有数据的特点和计量方法的适用性，创新性的构造了操作风险相关性度量的Common Shock模型，给出了一种操作风险相关性度量的方法。最后采用巴塞尔委员会2009年公布的操作风险损失数据进行了实证研究，定量分析了操作风险的相关性问题。对商业银行操作风险管理有一定的借鉴意义。
For the banking area, the operational risk measurement has last for several decades. And theoperational risk has been considered as one of the three most important risks banking confrontedwith credit risk and market risk. Analysis of the characteristics of operational risk, we can found ittaking place almost every day in banks’ operate. covered almost by all commercial bank businesslines and with great difficulty of measurement, by now still not a model can completely accuratemeasure of the size of the operational risk loss. In2004, Basel Ⅱ Accord the first time officiallydefined the operational risk and proposed three kinds of methods to measurement it, include BasicIndicator Approach, Standardized Approach and Advanced Measurement Approach. Use AdvancedMeasurement Approach to measure the operational risk is the trend of most of large commercialbanks, and for the most part of operational risk measurement research are still focus on it, becausebank uses this method could measure the operational risk on the basis of its own business feature,internal and external environment and risk management level to evaluate the risk conditionreasonable. So banks could count risk capital more accurate and prepare a good foundation for itssafety development, steady development and rapidly development.
In the process of use Advanced Measurement Approach to measure the operational risk, Theproblem of correlation has not been very good solved. The correlation questions can be understoodas a kind of contagion effect，Risk may happen or at the same time not happen，If the risk does notoccur in a very long time, So the commercial Banks may get to ignore the existence of this kind ofrisk. But, once it outbreaks, it very dangerous. So according to the correlation problem of operationrisk, This paper analysis the causes of the operating risk nature and connotation in all aspects ofsystem of the existing measurement methods on top, and then followed the introduction ofoperational risk’s correlation measurement. Point out the defects of use linear correlation coefficient.After that give the latest operation risk’s correlation measurement, common shock model andCopula function. Finally, use the operational risk loss data published in2009by the BaselCommittee to carry out a empirical research. Combined with correlation measure of CommonShock model, Given a method of measure the correlation between the operation risk, it will givesome reference to the measure the correlation between the operation risk In the commercial bank.
在操作风险的高级计量法当中，相关性的问题一直没有得到很好的解决。而相关性问题可以理解为是一种传染的效应，风险同时发生或者同时不发生。如果风险在很长时间之内同时不发生，那么商业银行可能会渐渐忽略此种风险的存在，逐步减少此项资本金的计提。但是，一旦风险爆发，很可能会严重威胁商业银行的生存与发展，造成无法弥补的损失。所以针对操作风险相关性的问题，本文首先在各个方面系统分析了操作风险的成因、内涵、特点、性质以及现有的各种度量方法，接着引入了操作风险相关性度量问题，指出目前使用线性相关关系度量相关性的缺陷，然后探讨了相关性问题研究的最新进展，给出两种相关性度量理论——Copula函数和Common Shock模型，并在对比分析两种相关性理论的基础之上，根据现有数据的特点和计量方法的适用性，创新性的构造了操作风险相关性度量的Common Shock模型，给出了一种操作风险相关性度量的方法。最后采用巴塞尔委员会2009年公布的操作风险损失数据进行了实证研究，定量分析了操作风险的相关性问题。对商业银行操作风险管理有一定的借鉴意义。
For the banking area, the operational risk measurement has last for several decades. And theoperational risk has been considered as one of the three most important risks banking confrontedwith credit risk and market risk. Analysis of the characteristics of operational risk, we can found ittaking place almost every day in banks’ operate. covered almost by all commercial bank businesslines and with great difficulty of measurement, by now still not a model can completely accuratemeasure of the size of the operational risk loss. In2004, Basel Ⅱ Accord the first time officiallydefined the operational risk and proposed three kinds of methods to measurement it, include BasicIndicator Approach, Standardized Approach and Advanced Measurement Approach. Use AdvancedMeasurement Approach to measure the operational risk is the trend of most of large commercialbanks, and for the most part of operational risk measurement research are still focus on it, becausebank uses this method could measure the operational risk on the basis of its own business feature,internal and external environment and risk management level to evaluate the risk conditionreasonable. So banks could count risk capital more accurate and prepare a good foundation for itssafety development, steady development and rapidly development.
In the process of use Advanced Measurement Approach to measure the operational risk, Theproblem of correlation has not been very good solved. The correlation questions can be understoodas a kind of contagion effect，Risk may happen or at the same time not happen，If the risk does notoccur in a very long time, So the commercial Banks may get to ignore the existence of this kind ofrisk. But, once it outbreaks, it very dangerous. So according to the correlation problem of operationrisk, This paper analysis the causes of the operating risk nature and connotation in all aspects ofsystem of the existing measurement methods on top, and then followed the introduction ofoperational risk’s correlation measurement. Point out the defects of use linear correlation coefficient.After that give the latest operation risk’s correlation measurement, common shock model andCopula function. Finally, use the operational risk loss data published in2009by the BaselCommittee to carry out a empirical research. Combined with correlation measure of CommonShock model, Given a method of measure the correlation between the operation risk, it will givesome reference to the measure the correlation between the operation risk In the commercial bank.
引文
①Basel Committee on Banking Supervision, International Convergence of Capital Measurement and Capital Standards: A RevisedFramework，Bank for International Settlements[J]. June2004.
②Halperin K.Balancing Act, Bank Systems and Technology [J].200122-25
③Nelsen R B. An introduction to Copulas[M].New York:Springer,1998
④Y. Malevergne&D. Sornette,2001."Multi-dimensional Rational Bubbles and fat tails: application of stochastic regressionequations to financial speculation," Quantitative Finance Papers cond-mat/0101371, arXiv.org.
⑤Gleen G, Meyers, The Common Shock Model for Correlated Insurance Losses, Variance1:1,[J].2007
【1】 Basel Committee on Banking Supervision, International Convergence of CapitalMeasurement and Capital Standards: A Revised Framework，Bank for InternationalSettlements[J]. June2004.
【2】 Bessis Risk Management in Banking，Chichester: Jone Wiley and Sons[J].2010
【3】 Cline W.R Financial Globalization,Economic Growth and the crisis of2007-09,Washington, DC[J].2010.
【4】 Junji Hiwatashi，HironshiAshida．Advancing Operational Risk Management Using JapaneseBanking Experiences[J].2002(11）:33-36.
【5】 Gleen G, Meyers, The Common Shock Model for Correlated Insurance Losses, Variance1:1,[J].2007
【6】Michael Power．The Invention of Operational Risk[M].The university of New South Wales，February14，[J].2003：3-4.
【7】Robert Hubner等编，李雪莲，万志宏译.金融机构操作风险新论[M].南开大学出版社，2005：137-249.
【8】Chapter15:FAA.Operational Risk Management[M].FAA System Safety Handbook,2000,(12):3-15.
【9】 Halperin K.Balancing Act, Bank Systems and Technology [J].200122-25
【10】Chernobai A Svetlozar R.T,Operational Risk:a Guide to BaselⅡCapital Requirements,Models and Analysis [J]. Hoboken,NJ: John Wiley2007
【11】Junji Hiwatashi, Hironshi Ashida. Advancing Operational Risk Management UsingJapanese,Banking Experiences [J].2002（11）：33-36.
【12】 Michael Power．The Invention of Operational Risk[M].The university of New SouthWales,February14，2003：3-4.
【13】 Hussain A Managing Operational Risk in financial Markets,[J].Oxford2000
【14】 Nelsen R B. An introduction to Copulas[M].New York:Springer,1998.
【15】Bouyé E,Durrleman V, Nikeghbali A,et al. Copula: an open fields for risk management.Working Paper of Financial Econometrics Research Centre[J]. City University BusinessSchool,London,2001.
【16】Embreehts.P,F.Lindskog,.Modelling Dependence with Copula’s and Applications to RiskManagement, Technical report,[J].Department of Mathematics ETHZ，2001
【17】 Genest,C.,K.Ghoudi “A semiparametric estimation proeedure of dependence parametersin multivariate families of distributions.”[J].Biometrlka82(3):2003，543-552
【18】 Romano,C..“Calibrating and simulating Copula functions:an application to the Italian stockmarket.”[J].Risk Management Function，Capitalia，Viale U.Tupini2002，180.
【19】 Fantazzini,D..Copula’s Conditional Dependence Measures for portfolio Management andValue at Risk,[J].Working Paper, University of Konstanz.，2003
【20】Patton, A.J..“On the Out-of-Sample Importance of Skewness and Asymmetrie Dependencefor Asset Alloeation.”[J].Journal of Financial Eeonometrics2(l):2004，130-168.
【21】 Alexander C O. Volatility and correlation methods, models and applications. RiskManagement and Analysis; Measuring and Modeling Financial Risk [J].(C. O. Alexander,Ed.)Wiley.2003
【22】 Nelsen R B.Dependence and order in families of Archimedean Copulas[J]. Journal ofMultivariate Analysis.1997,60:111-122.
【23】 Y. Malevergne&D. Sornette,"Multi-dimensional Rational Bubbles and fat tails:application of stochastic regression equations to financial speculation," QuantitativeFinance Papers cond-mat/0101371, arXiv.org.2001.
【24】 Cuculescu Ioan, Theodorescu Radu. Extreme value attractors for star unimodal Copulas[J].Comptes Rendus Mathematique.2002,334(8):689-692.
【25】沈沛龙、任若恩，新巴塞尔协议资本充足率计算方法剖析，金融研究，2002，(6)：22-31.
【26】巴曙松.巴塞尔新资本协议框架下的操作风险衡量与资本金约束[J].经济理论与经济管理，2003，(2):17-24.
【27】田玲，黄斌.保险在商业银行操作风险管理中的应用[J].投资科技，2003，(11)：143-145.
【28】樊欣，杨晓光.我国商业银行操作风险状况[J].管理评论，2005，(11)：43-47.
【29】王光升，赵昕，邵秀娟.商业银行操作风险量化的LDA与EVT综合方法研究[J].经济论坛，2005，(22)：94-96.
【30】曲绍强，王晓芳.我国利用损失分布法（LDA法）度量操作风险探析[J].经济问题，2006，(10)：65-67.
【31】李春光，丁浩.商业银行操作风险内涵分析[J].经济研究导刊，2009，(19)：58-60
【32】卢安文，任玉珑，唐浩阳.基于贝叶斯推断的操作风险度量模型研究[J].系统工程学报，2009，(3)：286-292
【33】王旭东．新巴塞尔资本协议与商业银行操作风险量化管理[J]．金融论坛，2004，(2)：1-6．
【34】王博，孟生旺.极值Copula在商业银行操作风险度量中的应用[J].统计教育，2010，（8）：55-60
【35】曹沫冰，我国商业银行操作风险的统计特征与防控对策[J].金融论坛，2009，(9)：21-25.
【36】王恒，罗刚.由低频高危型事件引发的操作风险的衡量方法浅析[J].科技信息，2006，(5)：55-56.
【37】陆静.运用贝叶斯网络方法构建操作风险预警系统—以零售银行业务为例[J].中国管理科学，2006，(14)：378-383.
【38】王朝晖.商业银行操作风险高级度量模型的设计与应用[J].华北金融，2006，(5)：21-25.
【39】孙彬、杨朝军、于静.基于Copula函数的国际证券市场传染效应实证分析[J].上海交通大学学报，2009，(4)：544-549.
【40】张磊，邹玲.在当前我国商业银行操作风险度量研究[J].当代财经，2008，(10)：59-62..
【41】史道济，姚庆祝.相关风险函数Var的界[J].系统工程，2009，(04)：11-15
【42】吴振翔,叶五一,缪柏其.基于Copula的外汇投资组合风险分析[J].中国管理科学，2004,(12):1-5
【43】陆静，张宏毅.基于损失分布模型的操作风险相关性研究[J].生产力研究，2007，（18）：40-47
【44】罗俊鹏,史道济.边缘分布和Copula对资产组合选择绩效的影响[J].统计与决策，2006(6):21-24.
【45】张尧庭.连接函数Copula技术与金融风险分析[J].统计研究,2002,(4):48-51.
【46】史道济,关静.沪深股市风险的相关性分析[J].统计研究,2003,(10):45-48.
【47】韦艳华,张世英.金融市场的相关性分析—Copula-GARCH模型及其应用[J].系统工程,2004,22(4):7-12.
【48】于波，陈希镇，华栋.基于蒙特卡洛法的国内股票市场的Copula分析[J].科学技术与工程，2008，(3)：1243-1248
【49】张杰，刘伟.基于时变Copula的股票市场相关性分析[J].商业经济，2010，(4）：66-69
②Halperin K.Balancing Act, Bank Systems and Technology [J].200122-25
③Nelsen R B. An introduction to Copulas[M].New York:Springer,1998
④Y. Malevergne&D. Sornette,2001."Multi-dimensional Rational Bubbles and fat tails: application of stochastic regressionequations to financial speculation," Quantitative Finance Papers cond-mat/0101371, arXiv.org.
⑤Gleen G, Meyers, The Common Shock Model for Correlated Insurance Losses, Variance1:1,[J].2007
【1】 Basel Committee on Banking Supervision, International Convergence of CapitalMeasurement and Capital Standards: A Revised Framework，Bank for InternationalSettlements[J]. June2004.
【2】 Bessis Risk Management in Banking，Chichester: Jone Wiley and Sons[J].2010
【3】 Cline W.R Financial Globalization,Economic Growth and the crisis of2007-09,Washington, DC[J].2010.
【4】 Junji Hiwatashi，HironshiAshida．Advancing Operational Risk Management Using JapaneseBanking Experiences[J].2002(11）:33-36.
【5】 Gleen G, Meyers, The Common Shock Model for Correlated Insurance Losses, Variance1:1,[J].2007
【6】Michael Power．The Invention of Operational Risk[M].The university of New South Wales，February14，[J].2003：3-4.
【7】Robert Hubner等编，李雪莲，万志宏译.金融机构操作风险新论[M].南开大学出版社，2005：137-249.
【8】Chapter15:FAA.Operational Risk Management[M].FAA System Safety Handbook,2000,(12):3-15.
【9】 Halperin K.Balancing Act, Bank Systems and Technology [J].200122-25
【10】Chernobai A Svetlozar R.T,Operational Risk:a Guide to BaselⅡCapital Requirements,Models and Analysis [J]. Hoboken,NJ: John Wiley2007
【11】Junji Hiwatashi, Hironshi Ashida. Advancing Operational Risk Management UsingJapanese,Banking Experiences [J].2002（11）：33-36.
【12】 Michael Power．The Invention of Operational Risk[M].The university of New SouthWales,February14，2003：3-4.
【13】 Hussain A Managing Operational Risk in financial Markets,[J].Oxford2000
【14】 Nelsen R B. An introduction to Copulas[M].New York:Springer,1998.
【15】Bouyé E,Durrleman V, Nikeghbali A,et al. Copula: an open fields for risk management.Working Paper of Financial Econometrics Research Centre[J]. City University BusinessSchool,London,2001.
【16】Embreehts.P,F.Lindskog,.Modelling Dependence with Copula’s and Applications to RiskManagement, Technical report,[J].Department of Mathematics ETHZ，2001
【17】 Genest,C.,K.Ghoudi “A semiparametric estimation proeedure of dependence parametersin multivariate families of distributions.”[J].Biometrlka82(3):2003，543-552
【18】 Romano,C..“Calibrating and simulating Copula functions:an application to the Italian stockmarket.”[J].Risk Management Function，Capitalia，Viale U.Tupini2002，180.
【19】 Fantazzini,D..Copula’s Conditional Dependence Measures for portfolio Management andValue at Risk,[J].Working Paper, University of Konstanz.，2003
【20】Patton, A.J..“On the Out-of-Sample Importance of Skewness and Asymmetrie Dependencefor Asset Alloeation.”[J].Journal of Financial Eeonometrics2(l):2004，130-168.
【21】 Alexander C O. Volatility and correlation methods, models and applications. RiskManagement and Analysis; Measuring and Modeling Financial Risk [J].(C. O. Alexander,Ed.)Wiley.2003
【22】 Nelsen R B.Dependence and order in families of Archimedean Copulas[J]. Journal ofMultivariate Analysis.1997,60:111-122.
【23】 Y. Malevergne&D. Sornette,"Multi-dimensional Rational Bubbles and fat tails:application of stochastic regression equations to financial speculation," QuantitativeFinance Papers cond-mat/0101371, arXiv.org.2001.
【24】 Cuculescu Ioan, Theodorescu Radu. Extreme value attractors for star unimodal Copulas[J].Comptes Rendus Mathematique.2002,334(8):689-692.
【25】沈沛龙、任若恩，新巴塞尔协议资本充足率计算方法剖析，金融研究，2002，(6)：22-31.
【26】巴曙松.巴塞尔新资本协议框架下的操作风险衡量与资本金约束[J].经济理论与经济管理，2003，(2):17-24.
【27】田玲，黄斌.保险在商业银行操作风险管理中的应用[J].投资科技，2003，(11)：143-145.
【28】樊欣，杨晓光.我国商业银行操作风险状况[J].管理评论，2005，(11)：43-47.
【29】王光升，赵昕，邵秀娟.商业银行操作风险量化的LDA与EVT综合方法研究[J].经济论坛，2005，(22)：94-96.
【30】曲绍强，王晓芳.我国利用损失分布法（LDA法）度量操作风险探析[J].经济问题，2006，(10)：65-67.
【31】李春光，丁浩.商业银行操作风险内涵分析[J].经济研究导刊，2009，(19)：58-60
【32】卢安文，任玉珑，唐浩阳.基于贝叶斯推断的操作风险度量模型研究[J].系统工程学报，2009，(3)：286-292
【33】王旭东．新巴塞尔资本协议与商业银行操作风险量化管理[J]．金融论坛，2004，(2)：1-6．
【34】王博，孟生旺.极值Copula在商业银行操作风险度量中的应用[J].统计教育，2010，（8）：55-60
【35】曹沫冰，我国商业银行操作风险的统计特征与防控对策[J].金融论坛，2009，(9)：21-25.
【36】王恒，罗刚.由低频高危型事件引发的操作风险的衡量方法浅析[J].科技信息，2006，(5)：55-56.
【37】陆静.运用贝叶斯网络方法构建操作风险预警系统—以零售银行业务为例[J].中国管理科学，2006，(14)：378-383.
【38】王朝晖.商业银行操作风险高级度量模型的设计与应用[J].华北金融，2006，(5)：21-25.
【39】孙彬、杨朝军、于静.基于Copula函数的国际证券市场传染效应实证分析[J].上海交通大学学报，2009，(4)：544-549.
【40】张磊，邹玲.在当前我国商业银行操作风险度量研究[J].当代财经，2008，(10)：59-62..
【41】史道济，姚庆祝.相关风险函数Var的界[J].系统工程，2009，(04)：11-15
【42】吴振翔,叶五一,缪柏其.基于Copula的外汇投资组合风险分析[J].中国管理科学，2004,(12):1-5
【43】陆静，张宏毅.基于损失分布模型的操作风险相关性研究[J].生产力研究，2007，（18）：40-47
【44】罗俊鹏,史道济.边缘分布和Copula对资产组合选择绩效的影响[J].统计与决策，2006(6):21-24.
【45】张尧庭.连接函数Copula技术与金融风险分析[J].统计研究,2002,(4):48-51.
【46】史道济,关静.沪深股市风险的相关性分析[J].统计研究,2003,(10):45-48.
【47】韦艳华,张世英.金融市场的相关性分析—Copula-GARCH模型及其应用[J].系统工程,2004,22(4):7-12.
【48】于波，陈希镇，华栋.基于蒙特卡洛法的国内股票市场的Copula分析[J].科学技术与工程，2008，(3)：1243-1248
【49】张杰，刘伟.基于时变Copula的股票市场相关性分析[J].商业经济，2010，(4）：66-69